How to Bend Wood


Over 30 strips of 2.5mm in place on the jig. This is just a dry run to check that every thing is ok.

There are many ways to bend wood but it will depend on the job that you want to do as to which method you will use. For example if you only want a slight bend, maybe simple brute force will be sufficient but it depends on the size of the wood. No amount of brute force will bend a foot square section of oak for example!

Thinner wood bends easier than thicker wood and some woods bend easier than others. The greater the curve, the thinner the wood needs to be. These are all basic rules that cannot be broken but if you know what is possible then wood can be shaped into almost any form. Even time can bend wood as can heat and moisture. Wood really is an amazing material.

As wood bends, the inner section compresses and the outer section expands. Sometimes it is possible to stop the wood breaking as it is bent by placing a steel band around it as it is bent into shape, in much the same way that a plumber inserts a steel spring into a copper pipe in order to bend it without it kinking.

Steam, or in fact any kind of heat can be used to bend wood but it helps if the wood is green, or in other words, freshly cut so it contains more moisture and is naturally more flexible. But steaming is only good for relatively thin sections of wood. So what does one do if you want to bend a larger section?

Well, it depends on how much you want to bend the wood. It also depends on the orientation of the grain and as I mentioned earlier on the type of wood. Often brute force and time can bend quite substantial pieces of wood but these days people are in a hurry and can’t wait months for gravity to do its thing.

Perhaps the most effective way to shape wood is to laminate it. This is done by gluing thin strips together using a jig to create the shape required. The thickness of the wood required will depend mainly on the radius of the curve required. The greater the curve, the thinner the strips will need to be.


Making the three sections of the jig from layers of 20mm plywood. This jig was made in three pieces because there are many knees to make and they are all slightly different. By keeping the curved section the same, only the other pieces need to be modified each time.

The down side of laminating wood is the fact that you have to make a jig and use glue. You can use almost any kind of glue but for the best and strongest results it’s best to use epoxy. The trouble with epoxy is that it is messy and expensive. That said, once the jig is made it’s easy enough to bang out any number of glued laminations.

Recently I was called in to reinforce a 20 metre wooden sailing boat. The deck had lifted due to the compression forces of the mast and rig and in order to put more strength into the boat new and larger hanging knees were required. So that the knees did not intrude into the boat too much they needed a fairly sharp radius but as they were 75 mm square that meant that they would need to be made from very thin strips indeed. In fact sometimes laminations can be made with wood as thin as one millimetre!


Just some of the strips of mahogany cut to make some laminated knees.

In this case it was decided that 2.5mm would be about right. The ideal is to use the thickest laminate possible as it requires less work and less glue and arguably looks better as there are less glue lines visible in the finished product. 2.5mm wood bends very easily even into a very tight curve but once you pile a load together it becomes much harder. It’s not so much the physical limits that you can bend the wood too but more the force needed to force a stack of laminations into the jig.

To create a 75mm thick knee, over 30 strips of wood were required and because there is quite a lot of force involved, the jig needs to be pretty strong. The jig was made from 4 pieces of 20mm plywood stacked together to create a support for the stack. The wood is actually cut about 8mm thicker than required as it will need planing down on both sides afterwards. When the wood is in the jig, it is slightly proud of the jig. This enables it to be bashed or clamped down into place.

The jig was made in three pieces and then they were bolted through the work table with 10mm bolts to ensure that nothing can move. Where the wood bends around the most, an inner block was also made to ensure that the clamping pressure on the curve was even.


The inner block is crucial to a quality lamination. This ensures that there are no visible glue lines in the finished product. The other areas are easier to clamp.

When the jig is made it’s time to do a dry run. This is to ensure that you will have no problems when you come to actually glue it up. What you do not want is to discover a problem half way through gluing. Epoxy is unforgiving in this respect. If it goes wrong, the wood will be glued like that forever and since epoxy, wood and your time is expensive, it’s worth taking a little while to do a dry run just in case. If the dry run is successful, remove the clamps and the strips and prepare some epoxy.

The stack is laid down on its side and the first strip is coated in glue. I always mix in a little thickening agent to ensure that the epoxy has a little flexibility and also to ensure that the epoxy goes on thick enough. Making sure you use the right amount of glue is a matter of experience but it’s certainly better to use too much glue rather than too little! Once the first strip is coated with glue (using a roller) the next is laid on top and then that is coated and so on until you get to the end.

Then the stack is placed into the jig which has been lined with plastic to avoid the assembly from gluing to it and thus making it impossible to remove once the glue has cured! The stack is clamped in place starting at the centre of the curve. As the first clamp is tightened a little, it’s worth bashing the stack down onto the work surface. Now work outwards in both directions clamping and bashing as you go.


The laminations, glued up and clamped in place. Leave to cure for 24 hours before removing.

If you have used sufficient glue and enough even pressure from the clamps you should end up with a solid piece of shaped wood that doesn’t even look like it was made from many pieces. Once it has been in the jig for 24 hours it can be removed. It is possible that the finished piece will ‘spring’ that means that it may open slightly as it tries to go back to its original shape. The more laminates you use, the less chance there is of the laminates springing.

Once out of the jig, an electric plane is the ideal tool for removing the glue that has hopefully squished out from between the strips. A belt sander also works well though sand with progressively finer grades to ensure that no scratches are visible in the finished product.


The knee, planed down and sanded. A bit of spirits on a rag show how the knee will look once varnished.

If you want to get really carried away you can assemble the stack for gluing so that the colour of the wood is uniform. Often wood, even from the same tree can be different depending on the direction or orientation of the grain so you can disguise the fact that a beam is made of many pieces by carefully matching each strip to the one above it. In reality it’s not really worth it unless you are trying to create something really spectacular. In time, most woods with either darken or lighten with exposure to sunlight which will make the colour more even.


A close up of the curved section of the knee. No glue joins are visible but it is possible to see a slight difference in colour of some of the laminations. In time these will fade slightly until you’ll have to look closely to see if the beam is made from many pieces.

You can of course make a feature out of using different woods in a lamination. Often boat tillers are made like this. It’s a bit cheesy but it does show what’s possible. Laminating wood is a great deal of fun and also very satisfying. It is also quite amazing how a stack of thin strips of wood that were collectively very floppy suddenly become unbelievably strong and resilient once glued.


Laminated tiller on a Dana 24. This one is made from mahogany and ash.


Simrad TP30 Tiller Pilot Long Term Review


Dave steers Doolittle mid Atlantic. Note his burgundy jacket, later changed for a beige one to keep him dry and cool.

Over the last couple of decades I have had a love hate relationship with autopilots. My last boat was a long keeled classic with a steep angle on the rudder and this caused no end of problems and over the years I tried and consequently destroyed most of the autopilots on the market.

By the time I had bought the Dana 24 I had ruined no less than 10 pilots and not one of them worked very well. When it came to buying an autopilot for ‘Doolittle’ I thought long and hard about buying an under deck unit but the way the Dana rudder is would have meant serious structural changes would have needed to be made. Then there’s the fact that under deck pilots are very complicated and consume a lot of power.

In the end it just didn’t seem worth it and besides, maybe the Dana with it’s vertical rudder and more modern shape would allow a modest tiller pilot to work. After a terrible experience with Autohelm (now Ratheon) I vowed I would never have anything to do with that company again so I went with the Simrad TP30.

Most autopilots make outrageous claims about the size of boat they can be used on. The TP30 is rated for boats up to 42 feet long so I figured that it should be about right for a 24 foot boat. Maybe if you are just using it in gentle conditions or motoring these claims are correct but I can’t imagine the TP30 steering a 42 foot yacht sailing in rough conditions.

Whether a tiller pilot works well on your boat will depend on how easy it is to steer when you are on the helm. If you have to put a lot of pressure or input in when sailing then the tillerpilot will struggle. If on the other hand, the boat is light on the helm and easy to steer (like most modern boats) the chances of an tillerpilot working are much greater.

On my last boat, the only way the pilot would hold a course was to set it up on it’s maximum gain (the speed of response to course change) and minimum sea state. The only trouble with this was that the tillerpilot was working like mad the whole time and this in itself caused problems. For some reason electronics manufacturers love to make everything black. I don’t know why they do this but when you are sailing in the Med in the summer, temperatures can get very high indeed. With a black horizontal surface soaking up the heat of the midday sun and the pilot working overtime, the pilot gets ridiculously hot. In fact one time in Sicily it got so hot that the cog spun on the motor shaft and rendered the pilot useless.

Suffice to say, I learned a lot about steering and tillerpilots over the years. The other thing I discovered is that the worst thing you can do, apart from using a tiller pilot on a classic yacht is to let them get wet. They don’t like it!

If you want your pilot to last the secret is this. Adjust it so it doesn’t work very much when steering the boat. By that I mean, do not load it up with excessive weather helm. If you are sailing you should reef the main very early so as not to load up the helm. Adjust the settings so that the boat can swing about a bit before the pilot starts steering. Don’t let it get wet or hot. I know that this is asking a lot but if you ignore this advice you will have all sorts of problems.


Pic shows the cog wheels. Two magnetic pins are pressed into the larger cog. If they slip out, they will smash the Hall sensors on the PCB.

The unit itself is mostly made of plastic and frankly isn’t very tough and there a few design faults which can render the pilot useless. The biggest problem are the magnetic pins in the larger cog wheel. These tell the pilot where the arm is and without them simply will not work. What normally happens is this: The magnetic pins start to work themselves out and then they smash into the delicate Hall sensors which are fitted to a small PCB. When that happens the pilot won’t work. It would be wise to always carry a spare PCB with the Hall sensors on it. It is easy to replace and very likely to get damaged sooner or later.

Another problem that you may experience is the slipping of the motor cog which is only pressed onto the motor shaft. Most of the time this cheap way of attaching the cog is sufficient but if the tillerpilot gets hot and is working hard it may very well start to slip and then the pilot will not work.

If you do buy a TP30 or the latest TP32 (which I assume is pretty much the same design inside) get used to the idea that you will have to take it apart on a regular basis. They claim that all you need to do is smear a bit of silicone grease on the shaft every now and then but this won’t be enough.


There are ten cross head screws in the underside of the pilot, unscrew them all and open the pilot like a book. (see pic). There is no need to dismantle the unit any more than this. The motor simply lifts out for easy access to it’s ends.

The most important thing you must do is push the magnetic pins out and glue them in properly with superglue. If you don’t do this, then one day the pins will ease out and smash the Hall sensors.

If you hear your pilot squeaking during use I recommend taking it apart and putting a drop of machine oil at either end of the motor itself. While the unit is apart put more grease on the worm drive and replace the belt if it is not tight. There is no tension device on the belt so if it is stretched it may slip.

You can take the panel off the electric part of the tiller pilot but this seems quite well sealed and in any case if the electrics do fail, there won’t be a lot you can do anyway so I suggest you leave it alone!

You can buy a spares kit which includes a new seal but I have found that if you don’t try to remove the seal when you open the unit it lasts for many years. In fact fitting a new seal is quite tricky so best not to touch it at all unless it is obviously broken.

To stop the tillerpilot over heating in the sun and to reduce the chance of water getting in make a jacket for it using a pale coloured material. Since our tillerpilot (Dave) has had a jacket he has been much more reliable. This is probably the most important thing you can do to keep your pilot working for as long as possible.

If you are having problems with the pilot suspect the electrical connections. If the pilot doesn’t get a good solid voltage it will cut out and behave erratically. Always check the socket, wiring and plug first before doing anything else.

So that covers the problems you are likely to have and how to avoid some of them. Let’s discuss the pilot itself.

At this point I have to say that we have had Dave on Doolittle for 6 years now and he has steered the boat across the Atlantic and has now done about 13,000 miles. This is most impressive for a plastic tillerpilot but there is no way he would have done this if not for his jacket and the regular oiling and greasing of his innards. He has had a new belt installed but apart from that he is original. The shaft is a bit pitted and corroded where some salt water has got in at some point but so far (touch wood) he is still working.

I think this says more about the Dana 24 than the quality of the TP30 in my opinion. The Dana is a very easy boat to steer and she will practically steer herself in most situations so Dave never has to work very hard and this must surely be the main reason why he is still working. That said, he has steered us bare poled before a gale mid Atlantic for 24 hrs and did an alright job.

The TP30 has an NMEA input so it can be made to steer to a waypoint but we always simply put the boat on a heading and press the red button. The controls are simple and the unit has coloured leds that light when you press a button. It also beeps so you know that the button has been pressed correctly. There are just 5 buttons to press for all the adjustments.

Dave can be used as a stand alone autopilot using his built in compass or you can fit a separate compass somewhere else on the boat. If you have wind instruments the unit can steer the boat by using the wind. There are also adjustments for gain and seastate.

There is an auto tack feature which when pressed tacks the boat through 100 degrees. Sadly this amount cannot be adjusted so when heading upwind I find that I have to head down 10 degrees or so before I press the tack button or the boat almost has the wind on the beam after I have tacked. It really is a shame that there is no simple way to adjust this to suit your boat’s best upwind angle.

It wasn’t cheap. At the time it cost about £450 which is a lot of money for a plastic tiller pilot but it does seem to work fairly well for most of the time. Basically the times it doesn’t want to work well are the times when if you took the helm you would have a hard time too. It’s quite powerful but not very fast nor is it predictive. It has to wait for the course to change before putting the helm over. To some degree the way you adjust the gain and sea state can help but the bottom line is that it has limitations.

No doubt a proper hydraulic under deck unit with separate compass and computers would be a much better way to steer the boat in more extreme conditions but as a compromise the TP30 isn’t bad. It is easy to fit. All I have to do is plug it in, drop the pin into the support and the other end snaps onto a pin on the tiller. It uses very little power and obviously the less it works the less power it will use but in any case I would be surprised if on Doolittle it takes more than 5 amps over 24 hrs which is brilliant.

Rather than drill a hole in my boat to fit it I simply added an Antenna mount bolted to the pushpit. It works very well.


A simple clamp on antenna mount is bolted to the stanchion and allows tillerpilot mounting without making holes in the boat. This has worked for over 13,000 miles so far.


If you have a modern boat that is easy to steer and light on the helm, the TP30 should work quite well.

It takes very little power and is quite powerful albeit rather slow to respond and move the helm.

It suffers a few design faults that can be easily sorted out. It also needs regular maintenance to keep it going.

It can be connected to NMEA, it can steer by separate compass or wind. There is a remote control as an option.

Don’t make it work hard and get hot and whatever you do, keep it away from water!